OpenSource P2P Debate, it's about choice

Well, here's the section of the GPL that I draw my conclusions from:


(Highlighting added by me of course.)

And, just for the record, here's the section I feel that prevents the makers of Moose from charging a fee of any sort:
 
This thread is going into a blatant Limewire commercial.

The technical things become mixed up, at the end it looks like Limewire never had or never lied about Spyware/Scumware, clustering of Limewire-only-clients is needed because of a over-complex less-efficient Superpeer concept (which Ultrapeers are IMHO), the commercial vendors do have a friendly politics and do not disadvantage other. Best rumours I heard this week. Gnutellaforums become as controlled as GDF is, I'm sorry that technical skilled developers do not speak up. Even if the pro speaker flood with more posts....

...you can't ignore Gnutella development is snail slow, inefficient and you can't ignore users and developers are unpleased about Bearshare/Limewire, which causing a spilt in Gnutella.

This thread reminds me so much to Microsoft. First every new competitor will be ignored, when it grows it will be badmouthed, then tryed to be blocked and finally flooded with wrong information and propaganda. Making windows non-modular so the MS midware is a must have, is like creating an unnecesary strange superpeer concept so you need to cluster your own clients (even old clip2 reflector showed you don't). Well, with a big enough market share you can try to dominate with such politics, but this is not the Gnutella idea. I learned from MS/Unix, fighting against comercial propaganda makes less sense... instead we have to show it can be done better, more efficient and with more technical inovations.

Join a open source project and fight the greed.
Do not cooperate with Bearshare and Limewire, make Gnutella a fair place.

Have fun, Moak
--
"As I would not be a slave, so I would not be a master. This expresses my idea of democracy." (Abraham Lincoln)
"Omnis enim res, quae dando non deficit, dum habetur et non datur, nondum habetur, quomodo habenda est." (Aurelius Augustinus)
 
Sorry,

but some people here (including me) are not very good friends from Vinnie, so they are "slightly" agressive in their view!

Did you ever visited bearshare.net? If not visit it and you will see what is really agressive!!!

But we respect your view and what you said - I do!!!

I want that Gnutella will stay Gnutella but I can
 
How? Prove your words!

Seperating is now taking away the freedom?

You should tell Vinnie that he takes away our freedom too!

Vinnie made the rules other follow the rules its simple!

Morgwen
 
There have been a lot of accusations and crap leveled at me for dropping 0.4 connection support but the simple fact is that the 0.6 handshake method is a MUCH more robust system for implementing "temp hosts".

In other words, when a 0.4 host connects to you and you are "full" (i.e. you have no free host slots) you have no choice but to send them 10 (or so) pongs and then disconnect.

During this time, the other host is free to send you queries and pings and pongs and what not, and you have no choice but to accept them (since shutting down the receiver would cause a TCP RESET and prevent the pongs from being delivered).

Using the 0.6 handshake, it is possible to inform the remote end that the servent is full, still deliver alternate IP addresses (via X-Try) and gracefully close the connection.

And its even possible to avoid a TIME_WAIT state, by simply waiting for the remote end to shut down (if you were the one who accepted the connection).

Clearly, this is WAY more efficient than supporting the old 0.4 connections. Personally, I *like* clean code, and it was very therapeutic to lop off the old 0.4 code since what was left is much cleaner, less buggy, and easier to work on.

It also performs much better, and allows us to leave the user with NO time wait states.

But there are certain people (called Trolls) who are just pissed off for one reason or another (admittedly I might have rubbed a few people the wrong way in the past), and see fit to criticize both BearShare and Vinnie (that's me) for dropping 0.4 connection support.
 
IANAL, but from the GPL:







If it's compiled into the gnucleus source code, and distributed together I believe you must also release the source code to these modifications. At least that's how I understand it. I still hope you're trolling (the gpl and this privacy crap? haha), or have a good explaination of how you interpret the gpl differently from me.
 
Before this I had no choice.
For example: it seems that when I connect to one BearShare node, then others connect back till I have nothing but BearShare nodes connected. (I am using Gnucleus)
I search and most of the results come from BearShare nodes.
I don't know about you, but BearShare people don't seem to know what good music is. I have my own tastes you know.
So now I can selectivly block BearShare or Morpheus, or even Gnucleus and move around the network to find my music or any other rare files.
It's so nice to have that choice!
If I hear something on the forums here about a client getting greedy, or spying on it's users, I have the choice to block that client and not support it by letting it use my computer network resources.
When I hear that that client has removed the spyware, I can simply change a few settings and that client is not on my bad list anymore.
Power to the people!
If I hear that a old client is creating a problem for the network, I can simply add it to my list to help the network without waiting for a developer group to decide what to do.
What could be so wrong with giving the users a choice?
 
In the past Limewire has installed spyware without asking - it still does - and it will do in the future, if users do not complain. No matter what nice words the PR is telling afterwards, investigate and check Limewire against spyware. Sooner or later Limewire has gone trough all existing spyware and scumware. Nice history so far! To keep yourself uptodate with what's missing:

http://www.simplythebest.net/info/spyware.html
http://www.computerevidencerecovery.com/startprog.htm
Brand new spyware, soon in a Limewire next to you!

Btw, someone should document which spyware has been installed from Limewire and Bearshare so far, which were without asking the users?
 
Here is an interesting analysis of some things which may be relevent. The correspondence about the bug report starts at the bottom of the page.

http://sourceforge.net/tracker/index.php?func=detail&aid=518918&group_id=4467&atid=104467

This doesn't really support anyone's views in particular, but illustrates that some of the technical issues are complex.

Nos
 
People have the choice to use any client they wish. Your attempts to diminish their quality of service is an abuse of the gnutella network.
 
??? I deny the lies of the past because I didn't lie about anything. It's like me accusing you of lying about something right now. What would you do??

Abused Gnutella clients? I still just don't get it. Stop playing the victim. It's an open protocol with open proposols that get approved by a standards body and implemented by those involved. If we were abusing other Gnutella clients, don't you think they'd mention it? We talk with them every day, and all we discuss on this issue is how to make UltraPeers work better ant the details of how they can implement them. Abuse? Not a peep. If there were abuse, we would have worked it out in the process of creating the UltraPeer protocols in the first place, because no one would have tolerated it.

Honestly, I just think you like the idea of having a little rebel network, and I think that's fine. Just don't go trying to bring down people who really don't deserve it. Just create an alternative network and leave whatever bitterness behind. Making interesting technology is the fun part, and I really thiink that's somehow gotten lost in all of this. I would honestly be really excited and impressed if you made a network that had really cool, innovative ideas in it that have not been implemented before because, again, that's the fun part.

Take care.
 
> Yes the MAIN problem is VINNIE but he is the chief coder so THIS IS OF COURSE A BEARSHARE ISSUE!
Even though Vinnie is a prime example of corporate style greed gone berzerk, it's not just him. Just wait to see what Morpheus does after a little while.
What happens when LimeWire goes under and/or gets purchased by a large corporation? It's all about profit.
Maybe the threat of blocking will keep them in line, I hope so.
The only new thing new here is you can now set blocking via the GUI and then blocking happens without you doing anything.
Before you could simply press "remove" for any client you wanted to remove from your connection list. And press "Stop" or even block IP's you didn't like in uploads.
I have seen people posting about doing this manually to try to move around the network to get a better "horizon". I am sure people used this method to block things they didn't like, or just for a power trip.
This is one feature you won't see in a pay for client!
I did mention that this new client works on Gnutella or the OpenSource P2P Net didn't I?
http://opensourcep2p.sf.net Stop the greed!
 
Ho Ho Ho Chi Minh!
Ho Ho Ho Chi Minh!

Next thing unregistered will do is sing:
"We shall overcome!"

I want something of whatever he's smoking!
 
Why should they have to switch in the first place? All Gnutella clients, regardless of which company developed it, needs to have complete and total access. I didn't want to bring up the KaZaA/Morpheus dispute, but there are certain similarities with these two debates.



You think it's bad to do what to users; serve them advertising or block their clients?



Oh, now this is interesting. Please don't tell me that the existence of the OpenSource p2p Net is the direct result of politics at play. Because Gnutella and politics should not mix. It's neither any of my business or interest to know what political agenda you subscribe to, but I don't think that it needs to play a role in the future of Gnutella. Gnutella is above all that; Gnutella is a statement of freedom. Let's keep it that way.



We need choice. Whether it be commercial or free, it's my view that the more clients we have, the better.

If more people would be using/supporting free clients, why are LimeWire and BearShare so popular? Correct me if I'm wrong, but LimeWire and BearShare, to my the best of my knowledge, have not been advertised extensively. LimeWire is a multi-platform client, which has led to it's popularity. BearShare locates more nodes faster, which has led to it's popularity. It's the program features, not just the marketing, that has led to each respective program success.

I also found it interesting that you mentioned free clients are being disadvantaged. As they are free, they are not in direct competition with anyone. They exist primarily to enrich file-sharing, and offer choice.


I disagree. The term OpenSource carries huge weight with it. It also carries responsibility.

You have termed the Network, not just the Client, as OpenSource. The Network is NOT Open. It is a private, members-only, exclusive network. Thus, this is a misuse of the term.

Yes, GPL is about the licensing of the program. But I didn't mention the term GPL. I'm talking about OpenSource. OpenSource reflects an image of freedom and equality. It's more than a license agreement, it's a manner in which we choose to conduct ourselves.



That doesn't mean you can use an OpenSource kiddy-hack script to shut down a store or businesses e-commerce server if you don't like what that business does. The argument you have just used entitles you to use your OpenSource word processor to create some Anti-Profit websites.


I don't buy it. It's called the OpenSource P2P Network. It's misleading.


Gnucleus offers a LAN edition. Check it out at http://www.gnucleus.net/general/download.html.
 
> This is so easy to prove that I'll let that as an exercise to Moak.

PS: Funny comment.

I thougt superpeers are designed to help also normal clients, e.g. to shield the weakest members (modem users). Thx Raphael, but I let you do your own homeworks and proof your clustering theories together with Vinnie and Limewire.

Actually I wonder about your flappy comment, Raphael. It is such easy to proof that clustering away of superpeers is not something you want to improve the network. Not all clients can act as superpeer (not enough bandwith, CPU or old OS), so not all Limewire users can act as superpeers, so you always get a mix of superpeer and normal clients. What do you wanna cluster, superpeers away from normal client? That makes no sense. As a matter of fact you will always have a mixture of superpeers and normal clients... exactly what was the idea behind superpeers, to balance and reduce load and traffic. So why or whom do you cluster? An exercise to Raphael.

That Limewire superpeer concept needs a clustering away from non Limewire clients to be reliable is one of the best rumours I heard in this propaganda circus, ridiculous. As I told before, maybe the Limewire proposal is not reliable enough and needs to be improved. All so far heared estimation are based on a Limewire _special_ superpeer concept and partly away from reality (e.g. a 500 times higher horizon without involving non-LW clients). The LW's proposal is unecesarry complex and inefficeint, there is no real need of clustering with a different concept (imagine a simple clip2 reflector but with improved eDonkey superpeer features and exchange of file databases). Perhaps in future you will find alternative concepts, without your self created clustering needs and without forcing others to do it like the commercial vendors do. In the past I was willing to work out solutions, but the high society GDF is not even intersted in listening and learning. The "Gnutella development community" was a very disapointing experience with less technical innovations the last months.

So, LW and BS are still valuable partners for you, good to know.
 
Back
Top